GB2480264A

GB2480264A - Telemetry apparatus and method for finding the most likely path taken by a vehicle

Info

Publication number: GB2480264A
Application number: GB1007801A
Authority: GB
Inventors: David K Linsdall
Original assignee: Thales Holdings UK PLC
Current assignee: Thales Holdings UK PLC
Priority date: 2010-05-10
Filing date: 2010-05-10
Publication date: 2011-11-16
Also published as: GB201007801D0; FR2959827A1; FR2959827B1

Abstract

Means of telemetry for recoding the paths taken by vehicles on a road or rail network comprising a processor and a data storage unit containing a map database. Data representative of the most likely path travelled by the vehicle on a map of the network represented in the map database is stored, and the processor is programmed to respond to an input identifying as a position fix the current position of the vehicle. The processor is further programmed to compare this position fix with the map database to determine one or more candidate geo-objects of the network map that are in the same area. The processor is also programmed, for each candidate geo-object, to find a possible path that the vehicle could has taken to reach that candidate geo-object from a geo-object that has previously been selected to be the best match to a previous position fix. The processor is also programmed to determine from those that have at least one possible path, the candidate geo-object that best matches the position fix, by taking into account quality parameters including the distance between the geo-object and the position fix.

Description

I

Telemetry Apparatus This invention relates to telemetry apparatus both for use in a vehicle and also for use in a back-office system, for recording the paths taken by vehicles, particularly on a road or rail network. The apparatus may use GNSS, Global Navigation Satellite System, or other positioning systems, and it is particularly useful for road vehicle user charging and for precise train positioning.

Electronic positioning systems, such as GPS, calculate a position (Fix) in a coordinate system that is referenced to a specific geodetic datum. In the case of GPS this is an ECEF system called WGS84 and provides Latitude, Longitude and Height over the specified spheroid. In most cases the user actually wants to know the position in relation to a physical object. This is particularly true when such systems are being used to position road or rail vehicles that generally travel on a fixed infra-structure. It is therefore necessary to convert the measured fix coordinates into another coordinate system based upon the identification of the location relative to the geographical object (geo-object) on the ground.

This process is traditionally called map matching and involves the best fit identification of the geo-object from the supplied general coordinates.

This is particularly true of road charging and rail signalling systems where the geo-objects are further used to determine the payment due or the control of the signalling system.

Although all map matching algorithms perform the translation from a geographical coordinate to a geo-object location there are many additional functions required and the actual choice of map matching algorithm is dependent upon the application.

The algorithm used for in-car satellite navigation has to provide a real time translation but has no interest in the past track since it is solely used to identify the start point of the future planned route. If a translation error is made, the system simply starts again with a new position.

Other published algorithms are designed to post-process an entire set of coordinates and determine the best estimate of the track followed by the vehicle.

The road user charging and rail signalling applications must combine both characteristics. The system must provide a real time estimate of the location and also ensure that the historical track is correct. The algorithm must also operate within the constraints imposed by the system. In particular, thin client systems separate the position measurement from the map matching in order to remove the problems of maintaining an accurate map database on thousands of vehicles. This means that the positions sent from the vehicle to the map matcher must be as infrequent as is possible so as to reduce the communications costs.

Accordingly, the present invention provides mobile telemetry apparatus for use in a vehicle travelling on a road or rail network, comprising a processor and a data storage unit containing a map database, the apparatus configured to store in a long term storage unit or to transmit to an external back-office system output data representative of the most likely path travelled by the vehicle on a map of the network represented in the map database, the processor being programmed to respond to an input identifying as a position fix the current position of the vehicle; and being further programmed (a) to compare this position fix with the map database to determine one or more candidate geo-objects of the network map that are in the same area; (b) for each candidate geo-object, to find a possible path that the vehicle could have taken to reach that candidate geo-object from a geo-object that the processor has previously selected to be the best match to a previous position fix; and (c) to determine from those that have at least one possible path, the candidate geo-object that best matches the position fix, by taking into account quality parameters including the distance between the geo-object and the position fix, and optionally also the heading of the geo-object relative to the fix heading and/or the overall distance travelled since the last fix.

The invention also provides telemetry apparatus for recording the paths taken by vehicles travelling on a road or rail network, comprising a processor and a data storage unit containing a map database, the apparatus configured to store data representative of the most likely path travelled by each vehicle on a map of the network represented in the map database, the processor being programmed to respond to an input identifying as a position fix the estimated current position of the vehicle, the processor further being programmed (a) to compare this position fix with the map database to determine one or more candidate geo-objects of the network map that are in the same area; (b) for each candidate geo-object, to find a possible path that the vehicle could have taken to reach that candidate geo-object from a geo-object that the processor has previously selected to be the best match to a previous position fix; and (c) to determine from those that have at least one possible path, the candidate geo-object that best matches the position fix, by taking into account quality parameters including the distance between the GEO-object and the position fix, and optionally also the heading of the geo-object relative to the fix heading and/or the overall distance travelled since the last fix.

The apparatus further provides a method of vehicle telemetry comprising receiving data identifying the estimated position of the vehicle as a position fix on a. map of a road or rail network, comparing this position fix with the map database to determine one or more candidate geo-objects of the network map that are in the same area; (b) for each candidate geo-object, finding a possible path that the vehicle could have taken to reach that candidate geo-object from a geo-object that has previously been selected to be the best match to a previous position fix; and (c) determining from those that have at least one possible path, the candidate geo-object that best matches the position fix, by taking into account quality parameters including the distance between the geo-object and/or the position fix, and optionally also the heading of the geo-object relative to the fix heading and/or the overall distance travelled since the last fix.

The invention is capable of accommodating the sparse data characteristics of thin client systems and it can provide a real time match to the geo-object and also ensure that the historical track, i.e. the path that has been taken by the vehicle, is correct even when the distance between measured positions spans multiple map segments i.e. geo-objects.

Preferably, the processor is programmed to (d) store for each position fix any other candidate geo-object and the possible paths to that candidate geo-object, and (e) in the event that a processor determines that there is no candidate geo-object with a possible path to the current position fix, to reject the previous position fix and to access the store to find the rejected candidate geo-object or objects that had had possible paths, associated with an earlier position fix; and (f) to repeat the said process of finding, for each current candidate geo-object, any possible paths to it from those rejected candidate geo-objects, and, if there is any possible path, to substitute that earlier position fix for the said previous position fix, to update the store to change the earlier position fix to the rejected candidate geo-object for which there was a possible path, and to repeat the aforesaid processes (b) and (c); but if there is still no possible path, then (g) to repeat the said processes (e) and (f) for a still earlier position fix; and (h) to repeat step (g) until a possible path has been found or else all earlier position fixes in the store have been accessed.

The iterative single path map matching processing of this preferred feature of the invention is particularly important for compensating for errors in position fixes, especially when the distance between measured positions spans multiple map segments.

In order that the invention may be better understood, a preferred embodiment will now be described, by way of example only, with reference to the accompanying drawings, in which: Fig. I is a block diagram of a preferred embodiment of the invention; Fig. 2 is a graph illustrating the calculation of parameters of a position fix relative to a geo-object; Fig. 3 is a portion of a map comprising several geo-objects connected together at nodes; and Figs. 4 and 5 show portions of a map and a selected vehicle path on that map, the selected path being corrected between that shown on Fig. 4 and that shown on Fig. 5.

A preferred embodiment of the invention is shown in Fig. 1, which is an on-board mobile telemetry apparatus for a vehicle travelling on a road or rail network. This apparatus communicates with a back-office system through for example a GPRS or WiFi communications channel. The network is represented by a map database, at least a portion of which is stored in the data storage unit of the on-board device shown in Fig. 1. The map database could be stored in the back-office system, and portions of it downloaded from time to time into the data storage unit of the mobile device.

Preferably, the database comprises representations of nodes and of segments linking those nodes, as is conventional. The mobile device has a processing unit which receives position data from a GNSS receiver. In this example, there is also a communications transceiver linking the processing unit with the back-office system, and a communications transceiver linking the processing unit with external road side equipment, for example to identify that the vehicle is passing between predetermined zones on the map. The processing unit accesses the data storage unit. A power supply is provided for all the components of the mobile telemetry device.

The processing unit is also connected to receive signals from an odometer in the vehicle, indicative of the distance travelled. Alternatively, the distance signal could be derived as a double integral of acceleration data measured by an on-board inertial measurement system. Alternatively, the distance signal could be derived from the integral of the differences in the position data measured by an on-board GNSS or other positioning receiver.

In this embodiment, it is the processing unit of the mobile telemetry device that determines the most likely path that has been taken by the vehicle between two known locations. However, in alternative configurations of the overall system, the computation may be shared with a processor in the back-office system, so that the mobile telemetry device would be a thin client, It is also feasible for the known position data to be sent to the back-office system and for the back-office system to compute the most likely path for that vehicle using the map database, in which case there would be no need for any portion of the map to be stored on board.

The most likely path that has been taken by the vehicle between two known locations at different times will now be described.

The inputs to the processing unit are a pair of basic segment matches (Selected and Target) comprising the unique identity of the segment plus the distance of the position along that segment. This is illustrated in the portion of the map of the network shown in Fig. 3. The algorithm executed in the processing unit finds a path through the road or rail network between the two segments that best matches an estimate of the actual distance travelled by the vehicle. This figure for actual distance is. derived from the distance signal described above. The discovered path with a length closest to the distance travelled is selected.

The algorithm works as follows: 1. If the selected segment is the same as the target segment there is no need to search for a path.

2. If the selected and target segments are different, assume that the path connects to the entry node of the selected segment according to the direction of travel and iterate through all the connections to that node to find a path.

3. If a path cannot be found then, if the selected segment allows bi-directional travel, try looking from the opposite end and iterate through all the connections to find a path.

4. Return the discovered path if found within the limits.

The iterate function may be implemented as a pair of nested logical loops which test all possible paths through the "tree" of connections.

At each node, where a segment joins other segments, the possible connections are determined, including any time dependent connectivity, from an estimate of the time at which the vehicle passed that point, and each connection is tested, by following all its connections, node by node, and so on until either the target segment is reached or the test limits are exceeded.

The test limits are designed to limit the search down any particular path as soon as it is evident that the path cannot reach the target. This significantly limits the number of iterations of the algorithm, especially when the segments are separated by a large number of nodes. Two tests are applied: a maximum node count and a maximum path length check.

The node count is supplied as a configuration parameter to the algorithm and is selected according to the characteristics of the vehicle sensor (e.g. maximum distance between reported fixes) and the granularity of the map database.

A maximum distance is calculated as the upper bound for the search allowing for underestimated distance travelled due to GPS errors and no actual vehicle odometer.

The algorithm that has been used effectively in real data trials is as follows: maxLengthLimit = 10.0 + (DistanceTravelled + EstimatedfixError) x Multiplier; Where Multiplier = 1.5 if the change in heading is less than 45 degrees or a non-GNSS distance signal is available, otherwise Multiplier = 2.0.

The algorithm checks the maximum path length, as each new node is tested, against the accumulated path length to that node plus the remaining direct distance from that node to the target fix position. Thus the test down a path will be terminated as soon as it is no longer possible to find a path within the maximum length limit. This test effectively constrains the search to a region approximately elliptical in shape with the test and target fixes at the foci, and so significantly limits the number of potential tests in a complex road network.

It is quite likely that a road network allows loops, since roads often connect to each other via several routes. These loops in the connection tree could lead to unnecessary tests following repeated segments so the algorithm checks if a segment has been seen earlier in the path and discards that particular "limb" of the tree.

The target segment is likely to be reached by several possible valid paths and so returns the path that best matches the target length parameter supplied.

Time dependent connectivity of nodes or segments is represented in the map database by data representing time intervals and linked to the nodes or segments. For example, a bridge or level crossing may break the network at time intervals that are recorded and are sent as updates of the database. As another example, a road may always be closed at predetermined times of the day.

In one embodiment of the invention the algorithm may be implemented as methods in a Java class that can be used from within a Java application or web service. The algorithm may process significant quantities of data from different measurement platforms and allows a very large number of vehicles to be processed in real time on a single server platform.

A preferred embodiment of the present invention will now be illustrated with reference to a specific map matching algorithm that uses the functions described above, to determine the best path through the network between GEO-objects, and applies these functions so as to determine the correct track and to manage the errors experienced with real data in obtaining position fixes.

The algorithm for the present invention is referred to as the iterative single path map matching algorithm, because, at any one time, it maintains a single estimate of a selected path through the network, but it can iteratively reprocess previous fixes, obtained earlier in time, and held in a state buffer in order to rectify an earlier incorrect selection that is discovered due to the processing of a subsequent position fix.

The iterative single path map matching algorithm operates as follows. The algorithm is part of a computer program that is executed in the processing unit of the mobile device shown in Fig. 1, or alternatively in a processor in the back-office system, or alternatively the processing is distributed.

First, for a given position fix that is input, its quality in terms of the vehicle heading and other parameters are validated. As shown in Fig. 2, the position fix is illustrated at the point marked Fix". Where the position fix is supplied by a GNSS system such as GPS, it will include the heading (angle of travel direction) of the vehicle, but it is well known that a measured GPS heading is only valid when the vehicle is moving above a threshold speed, typically 3 metres per second. Accordingly, this criterion is validated as a first step, to provide data representing the reliability of the heading measurement.

Next, the map database, of which at least a portion is stored in the data storage unit in the case of processing done in the mobile device, is accessed, and in particular an index of the geo-objects in the database is indexed, to obtain a candidate set of geo-objects that are in the same geographical area as the position fix. There may for example be several segments of the map in the geographical area, and these segments may not all be interconnected. For each candidate geo-object, the algorithm calculates the relationship between the geo-object and the position fix. With reference to Fig. 2, this relationship includes parameters such as the offset distance D and the distance L along the straight segment between the nodes P and Q. If the segment P Q of Fig. 2 were to be the selected path for the vehicle, then the assumption would be made that the position of the vehicle was as shown as the "Snap Point", a distance L along the segment. This would imply a positioning error of D. Next, the path discovery algorithm, described above with reference to Fig. 3, is used in order to find the best path consistent with the connectivity information contained in the map database, from each of the candidate geo-object matches back to the geo-object selection determined for the previous position fix, or back to an earlier selection if there are intervening fixes that could not be determined. Thus for each of the candidates for which there is a possible path, the path discovery algorithm selects the best match, according to a weighted combination of a predetermined set of the aforesaid quality parameters. Thus these quality parameters include the straight line distance from the geo-object, i.e. distance D in Fig. 2. They may also include the heading error, i.e. the information from the GNSS system indicative of the reliability of the heading measurement, as a function of the vehicle speed. The quality parameters may also include the difference in distance from the geo-object, between the present position fix and the last position fix.

In any case where there is no candidate geo-object for which there is a possible path that the vehicle could have taken from the previous position fix, the algorithm makes the assumption that an earlier selection of the path may have been in error and may therefore have led to an incorrect selection for the last position fix that was processed.

The data storage unit stores, for each position fix, the candidate geo-objects and the possible paths that have been computed to that candidate geo-object. This allows the algorithm to revert to stored earlier path selections, to review them and to correct the possible error. Accordingly, the processing unit scans back through the store, which is preferably a state buffer, until an earlier decision point is found, and it reprocesses earlier position fixes until either a path is found, or the end of the state buffer is reached. This scanning and reprocessing is described further below.

In any case where the solution cannot be determined, either the position fix is returned with no match, or the nearest candidate geo-object is selected, marked as the start of a new vehicle path. This selection may subsequently be shown to be the correct start of a subsequent path, or it may be shown to be bad (when subsequent fixes correctly link back to the earlier selections) due to a poor measured position, in which case it can be marked as a no match.

The reprocessing of earlier selections will now be described in greater detail. In the event that a new position fix cannot sensibly be matched to a geo-object with a path back to the previous position fix, it is necessary to review the matches selected for earlier position fixes, these matches being held in the state buffer, in order to determine if another choice would lead to a path at the current position fix.

The iterative single path algorithm takes the following steps: 1. It scans back through the state buffer until the first occurrence of the geo-object selection for the previous position fix. It is necessary for it to find only the first occurrence, if multiple position fixes are reported within the same geo-object as may occur for example at a road junction.

2. It then checks to see if there were any alternative selections, with valid paths, for that position fix.

3. It then checks to see if any of the candidate selections for the latest position fix being processed have a valid path back to this alternative, selection for the earlier position fix.

4. If there are any candidate selections with a valid path, then it changes the selection for the earlier position fix to its alternative selection, and it reprocesses the intermediate position fixes. It checks that the new selection that is made for each intermediate fix has both a path back to its predecessor and a path forwards to one of the candidates for the top level position fix..

5. If however no valid paths can be found at any of these points, then it repeats the process until all alternative selections have been tested for that previous position fix.

6. When all alternatives have been checked, then it repeats the process by looking further back in the state buffer until either a valid selection has been found, or the end of the state buffer has been reached.

Reprocessing of selections is illustrated in the map shown in Figs. 4 and 5. In Fig. 4, the selected path is shown in double lines. The position fixes are shown with crosses. The geo-objects, i.e. the segments joined at nodes, are shown in single solid lines, if they are not in the selected path. As shown in Fig. 4, a sequence of earlier position fixes ending with P0 had led the system to assume that the route taken was along path Ri. The subsequent three position fixes ending with position fix P1 had led the system to assume continuation of the path along the route R2.

Fig. 5 illustrates the same portion of the map, at the next position fix P2. As a result of this latest position fix, using the iterative single path processing, the route R3 is preferred to route RI, and it leads through path R4 to the current position fix P2. Accordingly, this has now become the preferred path, and the selected segments for the intermediate position fixes from P0 to P1 have been revised and stored accordingly.

U-Turns cause a particular problem in the road vehicle application in that a U-turn is a valid driving manoeuvre that needs to be detected, but unfortunately the position noise experienced by a typical GPS receiver at a junction in an urban environment can lead to false U-turns being detected with undesirable results. Experience during triats, where U-turns were included in the test routes to evaluate the ability to detect them correctly, led to a specific U-turn analysis and reprocessing function to be added to the ISP map matcher algorithm.

It is essential that the vehicle generate position fixes sufficiently often to allow U-turns to be correctly detected otherwise it is possible for U-turns to be removed or accepted incorrectly with a subsequent track error. The use of distance and direction change event triggers has been shown to be sufficient.

In one embodiment the function can be controlled by two configuration parameters: 1. Minimum significant distance travelled down a road prior to the U-turn 2. Percentage travelled down a road prior to a U-turn The iterative single path algorithm applies these tests if a road segment (geo-object) is exited via the same end as it was entered. A U-turn is only recognised if either the vehicle has travelled the larger of the distance or the percentage distance down the road segment. When an object is selected where the test fails, it is necessary to change the selection for that fix.

The algorithm takes all of the fixes in the state buffer that selected the same geo-object and makes a new selection of the nearest geo-object from either the one leading to that selection or that following the selection (i.e. the two geo-objects in the current path connecting to the entry end of the U-turn object).

In one embodiment of the invention the algorithm has been implemented as Java classes, together with supporting objects representing the position fixes, segment matches and paths, that are used from within a Java application or web service.

The Java classes are conveniently configured using an instance of a Java Properties object, passed to them when set up by the controlling application.

Configuration values are provided for all the key parameters to allow the algorithm to be tuned to the particular application where the geo-objects and position measuring devices may have varying characteristics.

In one embodiment of the invention the implementation is controlled by a state machine with the following five states: * Processing a new position fix for the first time * Processing the new position fix for the second or third time with different base parameters * Reprocessing the fix after reprocessing earlier fixes to correct an error * Reprocess an earlier fix by changing to an alternative candidate * Reprocess the fixes between a changed earlier fix selection and the current new fix due to an error having been corrected.

Claims

CLAIMS: 1. Mobile telemetry apparatus for use in a vehicle travelling on a road or rail network, comprising a processor and a data storage unit containing a map database, the apparatus configured to store in a long term storage unit or to transmit to an external back-office system output data representative of the most likely path travelled by the vehicle on a map of the network represented in the map database, the processor being programmed to respond to an input identifying as a position fix the current position of the vehicle; and being further programmed (a) to compare this position fix with the map database to determine one or more candidate geo-objects of the network map that are in the same area; (b) for each candidate geo-object, to find a possible path that the vehicle could have taken to reach that candidate geo-object from a geo-object that the processor has previously selected to be the best match to a previous position fix; and (c) to determine from those that have at least one possible path, the candidate geo-object that best matches the position fix, by taking into account quality parameters including the distance between the geo-object and the position fix.
2. Apparatus according to claim 1, in which the processor is programmed to (d) store for each position fix any other candidate geo-object and the possible paths to that candidate geo-object, and (e) in the event that a processor determines that there is no candidate geo-object with a possible path to the current position fix, to reject the previous position fix and to access the store to find the rejected candidate geo-object or objects that had had possible paths, associated with an earlier position fix; and (f) to repeat the said process of finding, for each current candidate geo-object, any possible paths to it from those rejected candidate geo-objects, and, if there is any possible path, to substitute that earlier position fix for the said previous position fix, to update the store to change the earlier position fix to the rejected candidate geo-object for which there was a possible path, and to repeat the processes (b) and (c) according to claim 1; but if there is still no possible path, then (g) to repeat the said processes (e) and (f) for a still earlier position fix; and (h) to repeat step (g) until a possible path has been found or else all earlier position fixes in the store have been accessed.
3. Apparatus according to claim 2, in which the processor is programmed in step (e) to access the store for the first occurrence of the geo-object selection for the earlier position fix, to take account of the possibility that there are plural position fixes with in the same geo-object.
4. Apparatus according to any preceding claim, in which the said quality parameters comprise the difference between the distances from the candidate geo-object to the current position fix and the previous position fix.
5. Apparatus according to any preceding claim7 responsive to a GNSS system output to locate the vehicle, and in which the said quality parameters comprise the possible error in a determination of the vehicle heading that is computed from the GNSS system data.
6. Apparatus according to any preceding claim, in which the processor is programmed to determine the most likely path to have been taken by the vehicle between two of its stored selected geo-objects, by determining the time actually taken to travel between them, and comparing that with the length of each possible path on the map between those two selected geo-objects.
7. Telemetry apparatus for recording the paths taken by vehicles travelling on a road or rail network, comprising a processor and a data storage unit containing a map database, the apparatus configured to store data representative of the most likely path travelled by each vehicle on a map of the network represented in the map database, the processor being programmed to respond to an input identifying as a position fix the estimated current position of the vehicle, the processor further being programmed (a) to compare this position fix with the map database to determine one or more candidate geo-objects of the network map that are in the same area; (b) for each candidate geo-object, to find a possible path that the vehicle could have taken to reach that candidate geo-object from a geo-object that the processor has previously selected to be the best match to a previous position fix; and (c) to determine from those that have at least one possible path, the candidate geo-object that best matches the position fix, by taking into account quality parameters including the distance between the geo-object and the position fix.
8. Telemetry apparatus according to claim 7, in which the processor is programmed to (d) store for each position fix any other candidate geo-object and the possible paths to that candidate geo-object, and (e) in the event that a processor determines that there is no candidate geo-object with a possible path to the current position fix, to reject the previous position fix and to access the store to find the rejected candidate geo-object or objects that had had possible paths, associated with an earlier position fix; and (f) to repeat the said process of finding, for each current candidate geo-object, any possible paths to it from those rejected candidate geo-objects, and, if there is any possible path, to substitute that earlier position fix for the said previous position fix, to update the store to change the earlier position fix to the rejected candidate geo-object for which there was a possible path, and to repeat the processes (b) and (c) according to claim 7; but if there is still no possible path, then (g) to repeat the said processes (e) and (f) for a still earlier position fix; and (h) to repeat step (g) until a possible path has been found or else all earlier position fixes in the store have been accessed.
9. A method of vehicle telemetry comprising receiving data identifying the estimated position of the vehicle as a position fix on a map of a road or rail network, comparing this position fix with the map database to determine one or more candidate geo-objects of the network map that are in the same area; (b) for each candidate geo-object, finding a possible path that the vehicle could have taken to reach that candidate geo-object from a geo-object that has previously been selected to be the best match to a previous position fix; and (c) to determining from those that have at least one possible path, the candidate geo-object that best matches the position fix, by taking into account quality parameters including the distance between the gèo-object and the position fix.
10. A method according to claim 9, comprising the further steps of (d) storing in a store, for each position fix, any other candidate geo-object and the possible paths to that candidate geo-object, and (e) in the event that it is determined that there is no candidate geo-object with a possible path to the current position fix, rejecting the previous position fix and accessing the store to find the rejected candidate geo-object or objects that had had possible paths, associated with an earlier position fix; and (f) repeating the said process of finding, for each current candidate geo-object, any possible paths to it from those rejected candidate geo-objects, and, if there is any possible path, substituting that earlier position fix for the said previous position fix, updating the store to change the earlier position fix to the rejected candidate geo-object for which there was a possible path, and repeating the aforesaid processes (b) and (c) according to claim 9; but if there is still no possible path, then (g) repeating the said processes (e) and (f) for a still earlier position fix; and (h) repeating step (g) until a possible path has been found or else all earlier position fixes in the store have been accessed.
11. A method according to claim 10, comprising in step (e) accessing the store for the first occurrence of the geo-object selection for the earlier position fix, to take account of the possibility that there are plural position fixes within the same geo-object.
12. A computer readable medium storing a computer program which, when executed by a computer causes the computer to carry out the method according to claim lOoril.
13. Mobile apparatus substantially as described herein with reference to the accompanying drawings.
14. Telemetry apparatus substantially as described herein with reference to the accompanying drawings.
15. A telemetry method substantially as described herein with reference to the accompanying drawings.